Home> Product List> MCP products > High Speed Gated Intensified Network Camera
High Speed Gated Intensified Network Camera
Model 303 Camera combines high-resolution digital network camera with extremely compact MCP image intensifier gating circuitry capable of exposures down to 10ns. “Network camera” means that it has 100 Mb Ethernet port and can be connected to any computer that has LAN hardware. It can also be plugged into an existent LAN directly or through regular hubs/switches or copper-to-fiberoptic converters if high EMI immunity is a must.
If synchronous operation of multiple cameras is required (as in a single multi-frame camera) they all can be networked together and input trigger should be applied in parallel to all cameras as each has it’s own programmable sequencer.
Software interfacing is also simple. Camera has embedded Linux processor running a web server - so any computer with Internet browser can access camera(s) to control it (them) and acquire images. No custom software applications/drivers are provided with camera - all user interface functions are performed by browsers interpreting web pages stored in the camera. This part of embedded software could be modified to meet specific requirements using only web design software and transferred to the camera by a standard FTP client.
Software running in the camera includes basic image processing - sensor per-pixel sensitivity calibration and correction. It fights both sensor’s fixed pattern noise and intensity modulation due to finite fiber size in MCP, it’s output window and fiberoptic taper. Additional application-specific software can be installed at customer request. All source codes of the software are available under GNU/GPL license.
The camera combines a fast gated micro-channel plate (MCP) image intensifier, a CMOS image sensor, and an embedded computer based on an Axis Communications ETRAX RISC processor running Embedded Linux. These three ingredients made it possible to create an extremely compact (1.5 x 1.5 x 5.5 in.), ultra-fast (10 ns exposure, +/- 0.8 ns jitter) camera module that does not require a frame grabber or special software to operate -- a standard 100Mbit Ethernet connection and a web browser are sufficient.
The Model 303 High Speed Gated Intensified Network Camera
Such cameras can be used individually if only a single snapshot of a fast event is
needed, or if the moving object is bright on a dark background so multiple-exposure
capability may be utilized. It also can be used as a module in a multi-channel camera.
That type of system consists of several cameras (usually one per frame), optical beam
splitter(s), and additional electronics to sequence cameras triggering and multiplexing
their outputs. When using Model 303 modules, standard Ethernet hubs and N-way input
trigger splitters can do the job, since each camera has it's own programmable sequencer.
Also, as a result, extending the data link to the camera and converting it to fiberoptic
is very simple and inexpensive, since regular copper-to-fiberoptic Ethernet converters are
used.
High Speed Gated Intensified Network Camera Questions and Answers:
What is a "gated intensified camera"?
This term refers to a class of ultra-fast cameras, with exposure times in the nanosecond
range. "Intensified" comes from image intensifiers -- vacuum tubes similar to
those for the nigh-vision devices. "Gated" means those tubes (in contrast to
night-vision applications) are used as shutters by applying fast electrical pulses to the
control electrodes.
Most such cameras can only make a single snapshot of an object, requiring that multiple
cameras (one per frame) be used when making even the shortest movies. This is due to two
reasons. First, advancing even a "digital film" or reading out information from
the image sensor takes time much longer than the exposure time. Even the fastest CMOS
sensors with small readout window settings are limited to tens of thousands frames per
second (fps), not a hundred million. The second reason is the long afterglow of the image
intensifier output screen, which is in the millisecond range for high efficiency
phosphors.
Why MCP?
MCP (micro-channel plate) image intensifiers use technology perfected for night vision
applications. They have gains in the thousands to tens of thousands range -- much higher
than any non-MCP intensifiers -- and gain is a very useful feature, because light is a
very scarce commodity with several-nanosecond exposures. There is, for example, a million
times less light in these exposure ranges as compared with regular still or video camera
applications with several-millisecond exposures.
Another important benefit of MCP image intensifiers is that they usually only need a
couple hundred volts to gate them, compared to the thousands needed to control other types
of intensifiers.
For basic information about microchannel plates click here.
Why CMOS sensor?
Most gated intensified cameras use CCD sensors, with fiberoptic tapers glued directly to
the chip surfaces. This provides more efficient coupling than lenses, if used for
transferring an MCP output image to the sensor surface.
CMOS image sensors (as compared to CCD) are usually "cameras-on-a-chip", with
all analog circuitry hidden inside, requiring just single low-voltage power supply,
accepting digital commands and generating digital output. Price for convenience -- lower
sensitivity and higher fixed-pattern noise -- pixel-to-pixel variation of sensitivity and
dark signal (pixel value when no light is applied). A detailed CCD/CMOS comparison is
available online at Dalsa, a manufacturer that produces both types of sensors. In this
particular sensor application where it is attached to the MCP image intensifier most of
the CMOS image sensor weaknesses are not significant.
Modern MCPs produce enough light for use with CMOS . "Enough" means that
increasing sensor sensitivity will not produce additional information (because of the
quantum nature of light detection on the MCP input) -- similar to trying to increase
contrast of a digital image when you already see brightness/color bands (steps) if the
original image is too dim.
The limited dynamic range of CMOS digital outputs (usually 8-10 bits) perfectly matches
(even slightly exceeds) that of the MCP.
The fixed-pattern noise of these sensors is masked by that of superposition of patterns of
micro-channels in MCP, fibers in its output (fiberoptic) window, fibers in the coupling
taper and pixels in a sensor chip. Note that fixed-pattern noise only looks nasty on the
raw images; it is not a real (random) noise, and may be eliminated without any information
loss by per-pixel calibration -- a simple operation for a camera that has sufficient
computational power and memory for storing an array of coefficients.
Are fast cameras required only for capturing fast events?
It is rather obvious that fast cameras are needed to capture fast phenomena -- capturing
explosions, ballistic tests, plasma studies, and lightning bolts.
There are also many other interesting applications not directly related to fast events,
such as very sensitive remote analysis using laser induced fluorescence (LIF) -- when
images are captured some time after the object was illuminated by a short laser pulse.
Another interesting application is that of looking through muddy water or just through
heavy snow at night when headlights do not help, Flash LADAR technology -- fast laser
illumination and gated image capturing is used to capture 3-d information.
It is possible, for example, to put on the same rotating platform a laser, gated camera
and second non-gated color one. Such a system is suitable for automatically building a 3-d
model of the scene around the camera -- the first camera is used to get the 3-d data, and
the second, apply textures.
| Del Mar Ventures 12595 Ruette Alliante #148 San Diego, CA 92130 v/fax: (509) 752-0123 mcp@sciner.com http://www.sciner.com/MCP/ |
 |
Copyright © 1999, 2000, 2001 [Del Mar Ventures]. All rights reserved.